Open data in the hotel industry: leveraging forthcoming events for hotel recommendation

The type, the nature, and the amount of information at the disposal of tourists have exploded, overwhelming them with the number of choices and making trip planning a very challenging task. Recommender systems offer personalized recommendations and help individuals overcome this information overload. The deployment of such systems in the hotel industry needs to satisfy specific constraints making the direct application of classical approaches insufficient. Travelers recurrently fall into the cold-start status due to the volatility of interest and the change in attitudes depending on the context. Events have been playing an essential role in drawing attention to specific regions and constitute a major motive to organize trips. In this paper, we explore for the first time the benefits of introducing open data related to events into hotel recommender systems. We address in particular two problems, the hotel-centric and the event-centric problems, where we recommend pairs of hotels and events in order to facilitate the tourism planning and enhance the travelers’ experience. We first collect data related to events and filter those susceptible of attracting people. We then develop a novel framework that infers events’ affinities based on hotel booking data and rely on the mass behavior to make personalized suggestions. Finally, we demonstrate our framework by performing a qualitative and a quantitative evaluation. Improvements in the quality of recommendation show the advantage of leveraging event data for hotel recommendation and illustrate the opportunities of using open data for tourism planning.